Achieving effective anticancer drug therapy requires achieving success in controlling at least three major physical properties: Toxicity within exposed cancer cells, delivery of the agent so as to exceed a threshold effective level of drug activity in the full anatomic extent of the cancer cell population, and limiting the side-effects affecting normal tissues. Recent preliminary data of our co-leaders indicates areas in which these critical barriers can be overcome. We found evidence for a severe limitation for effective drug penetration in pancreatic and perhaps other cancers. We developed new tools for a micro-anatomic understanding of drug delivery, methods to measure short-timed endpoints of drug effects, and ways to improve tumor-specific retention of delivered circulating agents. These developments create new opportunities addressed in our proposal. Our short-term Aims are to develop tools and understanding of the micro-anatomy and timing of drug penetration in GI cancers using patient and animal tumors, to manage better the toxicity of certain administered agents, and to improve drug delivery by and across the tumor vasculature. These studies are done in conjunction with four ongoing clinical trials (funded elsewhere) and may allow stratification of results from trials and the design of improved trials. Our long-term goal is to enable more efficacious treatment of GI cancers and to lay a foundation for using tissue-based drug-delivery assay as part of the routine lab accompaniments - the companion diagnostics - of cancer therapy.